Poster F101, Tuesday, March 27, 8:00-10:00 am, Exhibit Hall C
Correlation of memory regions with face and object regions differentially predict performance on face/object memory tasks
Michal Ramot1, Catherine Walsh1, Alex Martin1; 1National Institute of Mental Health, National Institutes of Health
There is great degree of variance in the population in face memory capabilities, with congenital prosopagnosia at one end, and “super-recognizers” on the other. In this study, we sought to characterize the networks underlying face memory, as measured by the Cambridge Face Memory Test (CFMT). To control for the specificity of the observed network for face / face memory processes, we also administered the Cambridge Car Memory Test, and the Glasgow Face Matching Test (GFMT) to investigate whether effects generalize across other face memory tasks. Healthy volunteers completed the three memory tasks outside the scanner, and then underwent a 3T fMRI scan, comprised of rest scans, a face/scene localizer, and passive viewing of a movie clip. We found that correlations during rest between the ventral face patches and memory related regions, specifically the hippocampus and surrounding parahippocampal tissue, strongly predict performance on the CFMT task performed outside the scanner (r=0.55), and to a lesser extent performance the GFMT. Correlations of the same parahippocampal region with object regions during rest predict performance on the CCMT. Correlations between the ventral face patches and face regions of somatosensory cortex were also strongly predictive of performance on the CFMT (r=0.52), but this somatosensory region did not predict performance on the CCMT. A similar but weaker pattern was found during the movie. These results suggest that performance on memory tasks has both general and domain specific substrates, and that the specific nature of the memory task affects the networks involved.
Topic Area: PERCEPTION & ACTION: Vision